Three Novel Mutations of the Spastin Gene in Chinese Patients With Hereditary Spastic Paraplegia FREE

Hereditary spastic paraplegia (HSP) comprises a group of inherited neurodegenerative disorders with the shared characteristics of slowly progressive spasticity and weakness of the lower limbs. Conventionally, they are divided into pure or complicated forms depending on whether the paraparesis exists in isolation or with other major clinical features, such as dementia, mental retardation, epilepsy, extrapyramidal disturbances, ataxia, deafness, retinopathy, optic neuropathy, peripheral neuropathy, and skin lesions.¹ Hereditary spastic paraplegia can be inherited in an autosomal dominant (AD), autosomal recessive, or X-linked manner.

To date, 19 loci (SPG1-SPG17, SPG19, and SPG20) that play a role in the development of HSP have been mapped. Eight genes for these loci have been identified.² The 4 dominant forms, at SPG3A, SPG4, SPG10, and SPG13, are the results of mutations in the genes that encode atlastin, spastin, kinesin heavy chain, and heat shock protein 60, respectively.^{3 - 6} The 2 recessive forms, at SPG7 and SPG20, are caused by mutations in the genes that encode paraplegin and spartin, respectively.^{7 - 8} The 2 X-linked forms, at SPG1 and SPG2, are the results of mutations in the genes that encode the L1 cell adhesion molecule and the proteolipid protein, respectively.^{9 - 10}

Forty percent of all AD families are linked to the AD locus (SPG4).⁴ The defective gene for SPG4 encodes a 616–amino acid protein named spastin. Spastin is an adenosine triphosphatase associated with various cellular activities (AAA) protein, with characteristic AAA cassettes in the C-terminal from amino acid 342 to amino acid 599. According to computer predictions, spastin possesses a nuclear localization signal, Walker motif A and B, an AAA minimal consensus sequence, leucine zipper motifs, and a helix-loop-helix domain. It is thought to act as a molecular chaperone and to play essential roles in many cellular activities, including cell cycle regulation, gene expression, vesicle-mediated protein transport, and protein degeneration.⁴ Recent evidence¹¹ indicates that spastin is distributed within the cytoplasm and interacts with microtubules. However, the function of spastin remains unclear.

There have been 124 different SPG4 mutations reported in North America, Europe, Japan, and Korea.^{4 ,12 - 29} To our knowledge, research on spastin mutations in Chinese patients with HSP has not been reported previously. To investigate the Chinese patients with HSP for mutations in SPG4, we analyzed SPG4 mutations in 31 apparently unrelated Chinese kindreds with HSP.

Thirty one unrelated Chinese kindreds with HSP were involved in this study. All probands were examined by a neurologist (B.T., G.Z., W.L., L.S., X.Z., or H.J.) and selected based on the Harding criterion³⁰ for the definition of clinical status, which was lower limb spasticity in the absence of any evidence of a structural lesion or demyelination. Disability was assessed on a 5-point scale: 1, normal or very slight stiffness in the legs; 2, moderate gait stiffness; 3, unable to run but able to walk alone; 4, able to walk with help; and 5, wheelchair bound.¹⁴ All patients were of Han nationality, from Hunan, Shandong, Heilongjiang, Zhejiang, and Guizhou in the People's Republic of China. Blood specimens were obtained from family members and control subjects after informed consent. Genomic DNA from peripheral blood leukocytes or lymphoblast lines was obtained by standard extraction methods.

According to the methods of Hentati et al,¹² the 17 coding exons of SPG4 (GenBank accession AJ246003) were amplified by the polymerase chain reaction using the original primer pairs. The reaction was performed in thermocyclers (Perkin-Elmer, Inc, Foster City, Calif), starting with an initial denaturation of 3 minutes at 94°C, followed by 30 cycles of 30 seconds' denaturation, 30 seconds' annealing at the primer-specific temperature, and 30 seconds' extension at 72°C. Electrophoresis of the polymerase chain reaction products was performed on 8% polyacrylamide (29:1 ratio of acrylamide-bisacrylamide) gels in 0.5% Tris borate EDTA, followed by silver staining. Samples showing altered migration patterns were selected for further analysis.

Abnormally migrating fragments were gel purified and automatically sequenced on a sequencer (model ABI 377; Applied Biosystems, Changsha, Hunan).

Variants identified by sequencing were analyzed for cosegregation in family members. All newly identified disease-associated variants were examined for presence in 50 controls (100 chromosomes) by single-strand conformation polymorphism analysis.

Comparison of means was performed using a 2-tailed t test and an analysis of variance test.

Of the 31 kindreds, 22 inherited as AD traits and 9 as sporadic. Patients in most cases presented first with hyperreflexia and spasticity of the lower limbs. The age at onset of symptoms showed a wide range, from 2 to 51 years. In the 31 probands, 22 were pure HSP and 9 were complicated HSP (characterized by spasticity and weakness of the lower limbs combined with cognitive deficit, ataxia, muscular atrophy, or ichthyosis) (Table 1).

Clinical features of the 4 AD families with SPG4 mutations are summarized in Table 2. Cranial nerves and power in the upper limbs were normal in all. In the lower limbs, spasticity in all was far more severe than any weakness present. There were only pure HSP patients in families H209 and H981, while pure and complicated HSP patients were in families H216 and H228. The age of onset in families H209 and H216 was much later than in the other 2 families. The mean ± SD age of onset in family H209 was 34.60 ± 6.69 years; in family H216, 34.50 ± 7.78 years; in family H228, 4.50 ± 5.00 years; and in family H981, 17.00 ± 2.83 years. The mean ± SD duration of the disease in family H209 was 15.80 ± 10.64 years; in family H216, 13.50 ± 9.19 years; in family H228, 25.25 ± 18.59 years; and in family H981, 29.50 ± 14.85 years. The son of the proband in family H981, aged 6 years, who was a mutation carrier, was asymptomatic and might be a presymptomatic patient, which could be due to age-dependent penetrance.

We detected 3 novel SPG4 mutations in 4 unrelated AD HSP families, including 2 missense mutations (T1258A and A1293G) and 1 deletion mutation (1668-1670delCTA) (Table 3 and Figure 1, Figure 2, and Figure 3). The mutation T1258A presented in 2 families simultaneously. All detected mutations were heterozygous and located in the highly conserved AAA cassette-encoding region of SPG4.

The mutation T1258A in exon 8 led to a replacement of leucine (a hydrophobic amino acid) with glutamine (a polar and hydrophilic amino acid) at position 378. The mutation A1293G in exon 8 replaced methionine (a polar and hydrophilic amino acid) at position 390 with valine (a hydrophobic amino acid). The 2 missense mutations occur close to Walker motif A, a conserved adenosine triphosphatase domain in the AAA cassette. The deletion mutation (1668-1670delCTA) in exon 14 resulted in a leucine deletion at position 515, which occurs in a leucine zipper motif. This is the second example of the deletion of an amino acid in spastin.²⁸ All variants were analyzed in 100 control chromosomes, and were not detected by single-strand conformation polymorphism analysis.

There was no significant difference between the average age at onset of all the probands with and without SPG4 mutations (F = 0.001, P = .98).

Three mutations of SPG4 were detected in 22 apparently unrelated Chinese AD HSP kindreds; to our knowledge, none have been described previously. In this study, SPG4 mutations accounted for 18% (4/22) of the mutations in AD HSP families, which is lower than the estimated 40% linked to this locus. The lower percentage may be correlated with racial differentiation and dissimilar genetic backgrounds because our methods are similar to those used in other SPG4 mutation analyses. No mutation (0/9) was detected in sporadic cases. Some sporadic cases could be de novo mutations of AD HSP, which would decrease the percentage of our families linked to SPG4. To our knowledge, this is the first report of SPG4 mutations in Chinese patients with HSP.

The 124 different SPG4 mutations reported previously are scattered along the coding region of the gene, except for exon 4, and include all types of DNA alterations, including 33 missense mutations (26.6%), 16 nonsense mutations (12.9%), 37 splice site mutations (29.8%), 10 insertions (8.1%), and 28 deletions (22.6%).^{4 ,12 - 29} The frequency of spastin splice site mutations was significantly higher than in patients with other human genetic disorders.^{14 - 15 ,21 - 23 ,29} A schematic diagram of SPG4 complementary DNA, including the position of the mutations identified in this study and in previous studies, is shown in Figure 4. The 3 novel mutations detected in the study and most of the previously reported mutations affected the AAA cassette of the spastin protein, which suggests this region is crucial for its function. All missense mutations, except 3, occur in the AAA cassette and could impair the spastin function through either a dominant negative effect¹¹ or a loss of function. The other types of mutations located inside or outside the AAA cassette-encoding region directly result in a stop codon or cause a frameshift with the subsequent introduction of a premature termination codon, which could lead to the production of a truncated protein lacking the adenosine triphosphatase domain. However, Charvin et al³¹ did not find a truncated protein when they performed protein analysis of lymphoblastoid cell lines of HSP patients carrying either nonsense or frameshift spastin mutations. Spastin transcript analysis provided strong evidence that transcripts were unstable in vivo, resulting in the absence or marked reduction of mutated spastin protein. These data strongly suggest that a dosage effect underlies the molecular mechanism.³¹

From all those previously described mutations, we find that most patients had unique SPG4 mutations, ie, recurring mutations are uncommon.^{4 ,14 - 16 ,20 ,23 - 24} We found a recurring mutation of T1258A in 2 families, suggesting that the mutation may have a higher frequency in the Chinese population, which could be helpful for future SPG4 diagnostic testing in the population. But the 2 kindreds came from the Hunan province, so we cannot exclude the possibility that a founder effect exists.

The number of different mutations found, together with the low yield of mutations in the small families and sporadic cases, suggests that mutation detection in spastin may not always by itself be a logical way forward for molecular diagnosis in patients with HSP. Future diagnosis of SPG4 mutations would benefit if we find some featured phenotypes associated with SPG4. Unfortunately, most patients in this study had a pure HSP phenotype with remarkable variation in age at onset and severity of the disease. For example, comparison of average age at onset between all the probands with and without SPG4 mutations did not reveal a significant difference (P = .98). Even in the same family, patients had phenotype variation, particularly age at onset.

In summary, 3 novel mutations we detected widened the SPG4 mutation spectrum, and we concluded that the percentage of SPG4 mutations in Chinese AD HSP patients is lower than estimated. The molecular mechanism underlying SPG4 should be further clarified, which would be helpful for the future treatment of HSP.

Corresponding author and reprints: Beisha Tang, MD, Department of Neurology, Xiangya Hospital, Central South University, Changsha, Hunan 410008, People's Republic of China (e-mail: bstang7398@yahoo.com.cn)

Accepted for publication September 17, 2003.

Author contributions: Study concept and design (Drs Tang, Zhao, Xia, Luo, Shen, Zi, and Jiang); acquisition of data (Drs Tang, Zhao, Xia, and Luo, Messrs Pan and Long, and Mrs Dai); analysis and interpretation of data (Drs Tang, Zhao, Xia, Luo, Shen, Zi, and Jiang); drafting of the manuscript (Drs Tang, Zhao, Xia, Luo, Shen, and Zi and Messrs Pan and Long); critical revision of the manuscript for important intellectual content (Drs Tang, Zhao, Xia, Luo, Shen, Zi, and Jiang, Messrs Pan and Long, and Mrs Dai); statistical expertise (Drs Zhao and Xia); obtained funding (Dr Tang); administrative, technical, and material support (Drs Tang, Zhao, Xia, Luo, and Shen, Messrs Pan and Long, and Mrs Dai); study supervision (Drs Zhao, Xia, Luo, Shen, Zi, and Jiang).

This study was supported by grant 2001AA227011 from the National 863 Projects, Beijing, People's Republic of China; and grants 30070273 and 39900047 from the National Natural Science Foundations of China, Beijing.

We thank all the families with HSP (and their referring physicians) for participating in the study; and Shu Wen, MSc, for critical reading of the manuscript.